Significance
Viral membrane fusion proteins or spikes, like those of influenza, paramyxo, and retroviruses, mature by furin cleavage in the infected cell into a form that can be activated by receptor binding and/or low pH. Although the cleavage of the precursor releases the fusion peptide at the end of the transmembrane subunit, structural studies have shown that this causes only a local change in spike organization. Here we have studied the effect of furin cleavage on the fusion protein of a γ-retrovirus, the murine leukemia virus, by cryoelectron microscopy. We found that this caused a major reorganization of the spike structure. This might explain the activation of the intersubunit disulfide isomerase, which is unique for the spike of these retroviruses.
Keywords: retrovirus, Env maturation, cryo-EM
Abstract
The trimeric Moloney murine leukemia virus Env protein matures by two proteolytic cleavages. First, furin cleaves the Env precursor into the surface (SU) and transmembrane (TM) subunits in the cell and then the viral protease cleaves the R-peptide from TM in new virus. Here we analyzed the structure of the furin precursor, by cryoelectron microscopy. We transfected 293T cells with a furin cleavage site provirus mutant, R466G/K468G, and produced the virus in the presence of amprenavir to also inhibit the R-peptide cleavage. Although Env incorporation into particles was inhibited, enough precursor could be isolated and analyzed by cryoelectron microscopy to yield a 3D structure at 22 Å resolution. This showed an open cage-like structure like that of the R-peptide precursor and the mature Env described before. However, the middle protrusion of the protomeric unit, so prominently pointing out from the side of the more mature forms of the Env, was absent. Instead, there was extra density in the top protrusion. This suggested that the C-terminal SU domain was associated alongside the receptor binding N-terminal SU domain in the furin precursor. This was supported by mapping with a SU C-terminal domain-specific antigen binding fragment. We concluded that furin cleavage not only separates the subunits and liberates the fusion peptide at the end of TM but also allows the C-terminal domain to relocate into a peripheral position. This conformational change might explain how the C-terminal domain of SU gains the potential to undergo disulfide isomerization, an event that facilitates membrane fusion.
Typically, the membrane fusing type I spike proteins of enveloped viruses are made in the rough endoplasmic reticulum (RER) as trimers of a precursor polypeptide (1). This is cleaved into a peripheral and a transmembrane subunit by furin or a related enzyme when the protein passes the Golgi complex en route to the cell surface for virus assembly by budding (2). The cleavage primes the spike protein for activation by the viral receptor and/or low pH to induce virus entry through fusion of the viral and the cell membranes. In particular, the cleavage releases the fusion peptide at the N terminus of the transmembrane subunit. With the fusion peptides, the transmembrane subunits can interact with the cell membrane and then bring together the cell and the viral membranes by back-folding their polypeptides into a six-helix bundle (3, 4). In the mature spike of the virus, this process is inhibited by the peripheral subunit, and it is only after the viral receptor and/or low pH has removed this clamp that the transmembrane subunit can function. The synthesis of the fusion protein in a precursor form prevents premature receptor and/or low pH-mediated activation in the virus-producing cell.
In the cases of influenza and paramyxovirus, the atomic structures of the precursor and mature forms of the membrane fusing proteins, the hemagglutinin, HA, and the F protein, respectively, have been determined (5–8). In both cases, the release of the fusion peptide by furin cleavage caused only a local structural change in the protein. In the present work, we have determined the structure of the membrane fusion protein (Env) precursor of a γ-retrovirus, the Moloney mouse leukemia virus (Mo-MLV), by cryo-electron microscopy (cryo-EM) and compared it with its more mature forms, which we determined before. We found that furin cleavage of the Env precursor resulted in a major change of the spike structure.
The peripheral or surface subunit (SU) of the Mo-MLV Env is composed of a receptor-binding N-terminal domain, RBD, which is linked to a C-terminal domain by a proline rich region (PRR) (9). Binding of the viral receptor, a cationic amino acid transporter, MCAT-1, to the RBD activates the TM subunits via the C-terminal domains (10−13). The latter are linked to the TM subunits by intersubunit disulfides, and these inhibit TM activation (14–16). In the C-terminal domain of SU, the Cysteine (Cys) residue of this bond is part of a CXXC motif, where the other Cys is free (14, 16). When the RBD activates the C-terminal domain, the thiol of this Cys becomes deprotonated and attacks the intersubunit disulfide, causing its isomerization to a disulfide within the motif instead. This gives the TM subunits the freedom to fuse the viral and cell membranes through refolding.
Cryo-EM analyses of the Mo-MLV Env structure have shown that it does not have the head and stem feature typical for the membrane fusion spikes of many other enveloped viruses, but an open cage-like appearance with tripod legs (17, 18). The protomeric unit of the Env, i.e., the SU−TM subunit pair, forms three protrusions, an upper, bent finger-like one, constituting the cage roof and the upper side, a middle one making a side lobe, and a bottom one constituting a spike leg. The atomic structure of the RBD has been fitted into the top, bent finger-like protrusion, the bottom one should represent the TM subunit, and the middle one, by deduction, the C-terminal domain of SU (18, 19). The Env can be activated by Ca2+ depletion and the activation process arrested at an intermediate, isomerization inhibited stage by alkylating the CXXC thiol (16). We have studied the structure of the isomerization arrested intermediate Env (IAS) and found that Env activation is initiated by the outward rotation of the RBDs (18). This opened a hole in the cage roof, which might be used by the TM subunits to reach the cell membrane with their fusion peptides.
The Mo-MLV Env has two precursor forms. Cleavage of the primary precursor, the furin precursor, by furin generates the R-peptide precursor. The latter still requires a second cleavage to prime it for receptor-induced activation. This is done by the viral protease, which in newly assembled virus cleaves off a 16-residue peptide, the R-peptide from the C-terminal, membrane-internal end of the TM subunits (20–22). In the infected cell, the R-peptide has been shown to inhibit cell−cell fusion of furin-cleaved Env (23, 24). We have inhibited the R-peptide cleavage with the protease inhibitor amprenavir and produced virus with the R-peptide precursor Env (25). Cryo-EM analysis of solubilized and isolated precursors showed that the SU part of Env, with the upper and middle protrusions, maintained a very similar structure as in the mature Env, but the TM legs, which were separated in the mature Env, were tied together at their lower end by the R-peptide (26). This should effectively inhibit a TM refolding mediated membrane fusion process. In this work, we have used a Mo-MLV mutant with furin cleavage site violations in Env and amprenavir to produce virus with the primary, furin precursor form of Env. Structural analyses by cryo-EM showed that the C-terminal domain of SU, with the disulfide isomerase, as mapped with a domain-specific antigen binding fragment (Fab), did not form the middle protrusion but was associated with the RBD at the upper part of Env instead. The implications of this major structural change by furin cleavage of the Mo-MLV Env precursor will be discussed.
Results
The Mutant 90-kDa Furin Precursor Enters into Particles and Shows Authentic Features.
The furin precursor Env has earlier been observed by SDS/PAGE in radioactive pulse chase experiments in Mo-MLV infected cells as an 80 kDa band (gp80), which matures, first by processing the N-linked sugar units and adding O-linked sugar units, to a 90-kDa form (gp90) and then by furin cleavage into the R-peptide precursor with disulfide-linked SU and TM subunits (27–30). We wanted to inhibit the furin cleavage by mutating the cleavage site and allow the 90-kDa form of the precursor to enter into virus particles, from which it should be easy to isolate as a trimer, after solubilization, by centrifugation in a density gradient and use for structural studies. However, several cleavage-deficient Env mutants have been constructed and, unfortunately, shown to be unable to enter into particles (31). When screening these earlier experimental results, we found that one mutant, R466G/K468G, did assemble into virus particles, albeit inefficiently. We constructed the corresponding proviral genome (pNCA-GG), produced [35S] Cys labeled virus in transfected 293T cells, isolated the particles (GG-virus), and analyzed them for Env by immunoprecipitation and subsequent nonreducing 13% SDS/PAGE. The results showed that Env was indeed incorporated into the particles (Fig. 1A, lane 1). The incorporation efficiency was about 10% of that observed for wt virus. Before using the mutant Env for structure analyses, we wanted to assure that it corresponded to the 90-kDa furin precursor of wt virus. The latter can be accumulated in wt virus-infected cells in the presence of a furin cleavage-inhibiting peptide (FIP) (32). This peptide is toxic to the cells and compromises protein synthesis. Therefore, it is not useful for furin precursor production, but can be used to follow the 90-kDa form of the furin precursor analytically by SDS/PAGE. In the presence of FIP, the 90-kDa furin precursor can clearly be seen in wt virus-infected cells together with the original 80-kDa form (Fig. 1A, lane 4). It is migrating somewhat slower than the 80-kDa furin precursor and, most importantly, identical to the Env of GG-virus (Fig. 1A, lane 1). We conclude that it is the 90-kDa form of the furin precursor that has been assembled into the mutant virus. This protein is also seen in pNCA-GG-transfected cells both in the presence and absence of FIP as expected (Fig. 1A, lanes 3 and 5). In contrast, in wt provirus-transfected cells, the 90-kDa form of the furin precursor is rapidly cleaved into the disulfide-linked SU−TM complex, which, as shown before, in nonreducing SDS/PAGE, migrates still slower than the 90-kDa furin precursor (Fig. 1A, lane 2) (32). However, in reducing SDS/PAGE, the latter complex separates into SU and TM (Pr15E) subunits (Fig. 1B, lane 2), not seen with the mutant furin precursor (Fig. 1B, lanes 1 and 3).
Fig. 1.
(A and B) The furin precursor mutant is incorporated into particles and migrates like the wt 90-kDa furin precursor (gp90) in SDS/PAGE. The 293T cells were transfected with plasmids encoding wt or GG-virus. At 24 h posttransfection, the cells were incubated in the absence or presence of 70 µM FIP for 1.5 h, pulse-labeled with [35S] Cys for 15 min, and chased for 40 min. The cells were solubilized and Env immunoprecipitated with αSU and separated on 13% SDS/PAGE under nonreducing (A, lanes 2–5) and reducing conditions (B, lanes 2 and 3). Furin precursor mutant from labeled virus was analyzed in parallel (A, lane 1 and B, lane 1). Note that lanes 4 and 5 in A are shown at high contrast to compensate for the lower protein synthesis in FIP-treated cells. (C) The furin precursor mutant has a cleavable R-peptide. [35S] Cys labeled wt or GG-virus was produced in the presence or absence of 0.8 µM amprenavir. The virus was solubilized and Env isolated by immunoprecipitation and deglycosylated with PNGase F. The samples were analyzed on reducing 7% SDS/PAGE. Shown are phophorimages of the dried gels. The viral proteins are indicated to the right and the migration of standard proteins to the left of each panel.
The furin precursor mutant described above was incorporated into particles in the presence of amprenavir to prevent R-peptide cleavage and maintain the primary form of the furin precursor. To find out whether the viral protease can remove the R-peptide from the furin precursor mutant and whether amprenavir inhibits the cleavage, we produced GG-virus in the presence and absence of 0.8 µg/mL amprenavir and analyzed immunoprecipitated Env by 7% reducing SDS/PAGE after deglycosylation with PNGaseF. The furin precursor mutant produced without amprenavir migrated slightly faster than that made with amprenavir (Fig. 1C, lanes 3 and 4). The latter band was also split into a major slower migrating one and a tiny amount of a faster one comigrating with the furin precursor produced without amprenavir. Both bands migrated considerably slower than the SU band of wt virus produced in the presence and absence of amprenavir (Fig. 1C, lanes 1 and 2). We conclude that the mutant furin precursor maintains a conformation that can be cleaved by the viral protease and that amprenavir inhibits the cleavage.
To further test the furin precursor mutant for authenticity, we analyzed whether it was possible to make it fusion-competent by artificial cleavage with trypsin. To this end, we incubated the furin precursor in GG-virus at several different concentrations of trypsin. We noted that it was much more sensitive than the mature Env in wt virus (Fig. 2A). This suggested structural differences between the precursor and the mature Env. However, we could not observe the formation of subunit-like digestion products at any condition. Nevertheless, we engineered an R-peptide less furin precursor provirus (pNCA-GGΔR), which we expressed in 293T cells, and tested the Env for trypsin-induced fusion with overlaid rat XC cells. Polykaryon formation did clearly take place using 12 µg/mL trypsin for 5 min at room temperature (Fig. 2D) in contrast to trypsin-treated mock-transfected 293T cultures (Fig. 2B) and transfected 293T cells incubated without trypsin (Fig. 2C). As expected, from the protein analyses of trypsin-treated GG-virus (Fig. 2A), the fusion efficiency was much lower than for wt provirus-transfected cells, which resulted in complete fusion of the mixed cell culture. We concluded that some fusion activity can be rescued from the furin precursor mutant with trypsin digestion.
Fig. 2.
Trypsin cleavage primes the furin precursor mutant for fusion activation. (A) Trypsin cleavage of furin precursor mutant. Labeled wt virus or GG-virus was incubated with 0–4 µg/mL trypsin for 30 min on ice. An excess of soybean trypsin inhibitor (SBTI) was added either before (+) or after (-) the incubation period. The virus was solubilized and Env immunoprecipitated and separated on reducing 13% SDS/PAGE. (B−D). XC cell−cell fusion. The 293T cells were transfected with pNCA-GGΔR (C and D) or mock transfected (B), overlaid with XC cells, incubated without (C) and with (B and D) trypsin as described (SI Appendix, Materials and Methods). Finally, the cells were incubated for 2 h at 37 °C to allow syncytia formation, fixed, and stained with Giemsa. Note the syncytia formed upon trypsin treatment in D.
Structure of the Furin Precursor Mutant Trimer.
GG-virus was produced in four 150 cm2 bottles with transfected 293T cells. The Env was solubilized in a mixture of 1M non-detergent sulfobetain (NDSB-201) and 0.3% Triton X-100 in the cold and affinity purified on a galanthus nivalis lectin column before isolation of Env trimers by sedimentation on a 0.4–1.2 M NDSB-201 gradient. Samples from fractions of the gradient were analyzed by Blue native (BN)-PAGE and silver staining. The mature trimeric Env control migrated between the thyroglobulin and ferritin marker proteins (Fig. 3A, lane 1). We observed a clear band migrating slightly faster than mature Env in the gradient fractions (Fig. 3A, lanes 3–5) and concluded that this represented the furin precursor trimer. The somewhat faster migration than the mature Env might be due to a more compact structure. Some slower migrating material contaminated the precursor in the bottom part of the gradient (Fig. 3A, lanes 5–7). The NDSB-201 was removed from fractions containing apparently pure trimeric furin precursor (fractions 6 and 7 in Fig. 3), and this Env, in 0.05% Triton X-100, was analyzed by cryo-EM. About 10,000 particle images were collected and processed to generate a 3D structure of the trimeric furin precursor mutant at 22 Å resolution. This is shown in Fig. 3B, where it is compared with that of the R-peptide precursor at corresponding resolution, which we determined before (26). Like the R-peptide precursor, the furin precursor displayed an open cage-like structure, but with the striking differences that the middle protrusion was missing and that the upper protrusion had become much bulkier. This was also evident when examining the activated intermediate form (IAS) of the furin precursor described below (see Fig. 5). The atomic structure of the RBD fits into the upper protrusion of the furin precursor, approximately as in the R-peptide precursor, but cannot account for all of its volume (SI Appendix, Fig. S1). Apparently, much of the C-terminal, i.e., the disulfide isomerase domain of SU, which represents the middle protrusion of the R-peptide precursor and the mature Env, is associated alongside the N-terminal domain of SU, the RBD, when the Env is made. Furin cleavage then allows it to relocate and obtain its peripheral location as the middle protrusion in the spike. This interpretation was supported by the mapping of the C-terminal SU domain in the furin precursor using the Fab of the rat monoclonal Ab 83A25.
Fig. 3.
Structure of the furin precursor. (A) Isolation of the furin precursor. galanthus nivalis lectin-purified furin precursor was subjected to sedimentation in a 0.4–1.2 M NDSB-201 gradient. Samples from gradient fractions were analyzed for trimeric precursor by BN-PAGE and silver staining. Mature Env trimers (Mature gp3) and furin precursor trimers (Fur prec gp3) are indicated to the right and the migration of standard proteins to the left. Material in fractions 6 and 7 were used for cryo-EM. Note the presence of some slower migrating material (X) in fractions 8 and 9. (B) A comparison of the surface-rendered 3D reconstructions of the furin precursor structure (Fur prec) with that of the R-precursor (R prec) as determined before. Two side views and one top and one bottom view are shown. One protomeric unit is hatched in side view 2, and characteristic protrusions are indicated in side view 1. Number of particles in data sets and resolutions achieved are indicated. Scale bar represents 50 Å.
This monoclonal Ab reacts with most MLV strains, including Mo-MLV, and is directed against an epitope involving the most downstream small disulfide loop of SU, about 30 residues from the furin cleavage site (33, 34). The Ab inhibits Env activation at a step after receptor binding (35). The mAb immunoprecipitated the mature and both precursor forms of Env with equal efficiencies. We made complexes between the furin precursor and the 83A25 Fab in GG-virus and isolated the Env-Fab complexes as described above. However, although we incubated the virus with a molar excess of Fab, the BN-PAGE analysis of the final gradient sedimentation showed that the Fab occupancy on the trimer did not reach completeness. Nevertheless, we used the material for negative-stain EM and made an effort to process particle images slightly larger than unliganded Env trimers. The reconstructed structure showed the furin precursor with additional density on one side of the upper protrusion (Fig. 4, Center). The latter must represent the Fab binding to the Env, although a complete Fab could not be reconstructed, possibly due to flexibility of the Fab molecule and/or substoichiometric ligand binding to Env. Indeeed, the additional density corresponds roughly to the volume of the variable fragment of an Ab (Fig. 4, Right). We concluded that the enlarged upper protrusion in the furin precursor represents the RBD in association with the C-terminal domain of SU.
Fig. 4.

Structure of the furin precursor in complex with the 83A25 Fab. The Fab was complexed with the furin precursor in GG-virus by incubation for 16 h in the cold, solubilized and isolated by lectin chromatography and sedimentation in an NDSB-201 gradient, and analyzed by negative-stain EM. Shown is a surface-rendered structure reconstruction of uranyl acetate-stained Fab liganded furin precursor in side view 1 (Center). For comparison, the corresponding view of (Left) the unliganded precursor (from Fig. 3) and (Right) the liganded furin precursor, with the atomic structure of the variable fragment (cartoon representation) of the monoclonal Ab 7E2 against the Paracoccus denitrificans cytochrome c oxidase fitted into its additional density, are also shown.
To investigate the possibility that the characteristic features of the furin precursor were caused, not by the cleavage inhibition, but by the GG mutation at the cleavage site, we also isolated a hemagglutinin (HA) epitope-tagged 80-kDa form of the furin precursor (gp80) from provirus-transfected cells for structural analysis by negative-stain EM. The analysis showed that, like in the gp90 form, the middle protrusion was truncated and the upper protrusion was bulkier than in the furin-cleaved R-peptide precursor (SI Appendix, Figs. S2 and S3). However, the upper protrusion was differently shaped in the gp80 than in the gp90 furin precursor. All together, the results support a furin cleavage-mediated Env reorganization.
The Furin Precursor Undergoes Initial Steps of Activation.
We showed before, using biochemical experiments, that the furin precursor contained the disulfide that shackles the SU and TM subunits in mature Env (32). Furthermore, we demonstrated that the precursor could be activated in vitro by Ca2+ depletion, like the mature Env, to isomerize this disulfide. However, the isomerization activation was significantly impeded compared with that of the mature Env. In the latter form, as well as in the R-peptide precursor, the isomerization of the intersubunit disulfide is associated with an opening of the Env cage roof through an outward turning of the RBDs and we wondered if this was also the case with the furin precursor despite the complex between the RBD and the C-terminal SU domain (18, 26). Therefore, GG-virus was solubilized in the presence of EDTA and the thiol alkylating agent N-ethylmaleimide to convert the 90-kDa furin precursor into the activated intermediate state, IAS, and isolated for cryo-EM. The reconstruction of the 3D structure of the IAS furin precursor clearly showed the association of the C-terminal SU domain with the RBD, like in its native state (Fig. 5, Left, side views). However, despite this association, the whole domain complex opened the cage roof of Env through an apparent outward rotation, like the RBDs did upon activation of the R-peptide precursor and the mature Env (Fig. 5, Left, top view). We concluded that the furin precursor can undergo initial activation steps.
Fig. 5.
Structure of the furin precursor in its activated intermediate form, IAS. GG-virus was solubilized in the presence of EDTA and alkylator to generate the IAS form of the furin precursor. This was isolated by lectin chromatography and gradient centrifugation for structure reconstruction by cryo-EM. Shown is a comparison of the surface-rendered 3D reconstruction of the IAS furin precursor structure (IAS Fur prec, Left) with that of the IAS form of the R-peptide precursor Env (IAS R prec, Right) as determined before.
Discussion
We have here studied the structure of the furin precursor of Mo-MLV, i.e., the form of Env neither processed by furin into SU and TM nor by the viral protease to remove the R-peptide. The precursor was generated by violating the furin cleavage site by mutation, R466G/K468G, and by inhibiting the viral protease with amprenavir. The mutant was expressed as part of the corresponding provirus in transfected 293T cells and shown to assemble, although inefficiently, into particles from which it could be isolated for cryo-EM studies. The analyses showed that the middle protrusion, so characteristic for the furin-cleaved forms of Env, was missing and that the upper protrusion was much bulkier. Several properties of the mutant precursor suggested that it had an authentic structure. Like the wt precursor, it matured, by N-inked sugar processing and O-linked glycosylation in cells from an 80-kDa to a 90-kDa form and it oligomerized into trimeric complexes. Unlike the wt precursor, which rapidly was cleaved by furin into disulfide-linked SU-TM complexes, the mutant accumulated in the 90-kDa form. Although inhibited in furin cleavage, the R-peptide could be removed by the viral protease in particles when the amprenavir was omitted. Most importantly, it was possible to rescue some receptor-induced cell−cell fusion activity (fusion from within) with trypsin incubation, which apparently generated some functional subunits from the mutant precursor at the cell surface. Earlier studies have also demonstrated that an MLV vector with a similar Env mutation, R466A/K468A, was, when expressed in cells, able to interfere with the superinfection of the cells by another MLV vector using the same receptor (31). The mutant Env most likely bound to and down-regulated the receptor in the cells.
We also studied the structure of an HA-tagged 80-kDa form of the furin precursor by negative-stain EM and found that it, like the 90-kDa form, was lacking the middle protrusion and carried a bulkier upper protrusion than the furin-cleaved forms of Env. However, the upper protrusion was differently shaped in the 80-kDa and the 90-kDa forms. This could be a consequence of the glycosylation difference between the two forms. Also, the possible effects of the acidic HA nanopeptide insertion into the PRR SU region of the 80-kDa precursor and the limited resolution of the latter structure must be considered.
As the middle protrusion in the furin-cleaved forms of Env should represent the C-terminal SU domain, we interpreted the furin precursor structure to mean that much of this domain was associated with the RBD in the upper protrusion making it bulkier. This topology was supported by the cryo-EM mapping with the C-terminal domain specific Fab of the mAb 83A25 and by the RBD fitting. The association of the C-terminal domain with the RBD might be mediated by the N-terminal part of TM, possibly the fusion peptide and/or its flanking region, because furin cleavage just upstream of the fusion peptide in the precursor apparently allows the C-terminal domain to relocate into the middle protrusion position. After furin cleavage in wt Env, one might speculate that the fusion peptide remains in complex with the RBD. Here it could stay protected by the tips of the bent finger-like RBD until Env is activated by receptor binding. Then, the outward rotation of the RBDs could release the fusion peptides into the opening in the Env roof that the RBD motion creates. Interestingly, mutations in RBD have been described that inhibit most of the cleavage of the furin precursor into SU and TM, but not its transport to the cell surface and incorporation into virus (36). Possibly, these mutations distort the cleavage site via the putative RBD−fusion peptide interaction. This model also predicts that the RBD and the C-terminal SU domain of the furin precursor assemble first when the N-terminal part of the TM ectodomain with the fusion peptide has been synthesized in the RER of the infected cell. This assembly event might be guided by the formation of the intersubunit disulfide, linking the small disulfide loop in the middle of the TM ectodomain to the CXXC motif in the N-terminal part of the C-terminal SU domain (14, 37). Just upstream of the CXXC motif, there is a conserved and essential N-linked sugar unit that might be important to mediate binding to folding chaperones for this assembly step (38–41). The newly assembled unit with the RBD, the TM N-terminal region, and the SU C-terminal domain might, furthermore, maintain necessary surface structures for trimerization of the Env precursor protomers.
The reorganization of the Env structure that we found in the Mo-MLV upon furin cleavage distinguishes the Mo-MLV Env from other type I membrane fusion proteins, where this maturation step does not cause major structural changes. For instance, in the case of the parainfluenza virus 5 fusion protein (F), there is only a minimal structural change upon cleavage (7). Furin cleavage of the respiratory syncytial virus fusion protein results in a significant movement of the C terminus of F2 and the N terminus of F1 apart from each other, but the structural change remains local (8). A similar local change is seen with influenza hemagglutinin when it converts from the precursor to the mature form (5, 6). Even in the case of HIV-1, there seem to be no drastic changes in the overall structure of the spike protein upon furin cleavage (42, 43). The reorganization seen with the Mo-MLV Env might activate the disulfide isomerase in the C-terminal SU domain. We found earlier that the intersubunit disulfide isomerization activity was present in the furin precursor, but this was increased eightfold by furin cleavage (32). A corresponding structural reorganization might be present in other γ-retroviruses and also in the δ-retroviruses, like human T-cell leukemia virus, which also possess an isomerase domain (44).
With this work, we complete a structural analysis of protein domain movements in the Mo-MLV during maturation and activation using cryo-EM and single-particle image processing. These include, in addition to the furin precursor, the R-peptide precursor, the mature Env, and the activated intermediated form of Env, the IAS (18, 26). The structure of the final activated form, the trimer of back-folded TM, has been determined at atomic resolution earlier (3). We found that the three protein domains of the protomeric unit of Env, the RBD, the C-terminal domain of SU, and the TM domain, undergo major reorganizations during the maturation/activation steps although the overall cage-like, open spike structure is maintained (Fig. 6). In the furin precursor, much of the C-terminal SU domain is associated alongside the RBD (Fig. 6, Fur prec). After furin cleavage, it relocates, forming the middle protrusion of the R-peptide precursor (Fig. 6, R prec). Cleavage of the R-peptide releases threefold interactions of the TM subunits, allowing them to obtain a peripheral location in the mature Env (Fig. 6, Mature). Activation of the Env results initially in an outward rotation of the RBDs resulting in the opening of the spike roof (Fig. 6, IAS mature). This passage is probably meant for the extended form of the TM subunits, but this we have not yet observed, and it constitutes a major future challenge of research.
Fig. 6.
Structural changes of the Mo-MLV Env during its maturation and activation. Shown are the surface-rendered 3D reconstructions of the furin precursor (Fur prec), the R-peptide precursor (R prec) (26), the mature Env (Mature), and the Env at the intersubunit disulfide isomerization arrested stage (IAS Mature) (18), in side, top, and bottom views.
Materials and Methods
The Mo-MLV provirus mutant (pNCA-GG) encoding a furin cleavage-deficient Env was constructed by PCR mutagenesis. The corresponding virus was produced in transfected 293T cells and the 90-kDa form of the furin precursor purified, after solubilization with Triton X-100, by lectin chromatography and sedimentation in a NDSB-201 gradient. The activated intermediate state of the furin precursor was induced by Ca2+ depletion as described (26). The provirus encoding a cleavage-deficient 80-kDa form of the furin precursor with an inserted HA tag was also constructed by PCR mutagenesis, but the Env was purified from transfected cells using antibody and lectin affinity chromatography and subsequent sedimentation in a sucrose gradient. Cryo and negative-stain EM data were collected on a JEOL 2100F transmission electron microscope for reconstruction of the 3D structure by single particle image processing. Methodological details are given in SI Appendix, Materials and Methods.
Supplementary Material
Acknowledgments
We thank Maria Ekström for 83A25 Ab production and Martin Lindahl for help with tilt-pair analysis. Swedish Science Foundation Grant 2778 and Swedish Cancer Foundation Grant 0525 (to H.G.) and the EU FP7-People-ITN-2008 Marie Curie Actions Project Virus Entry 235649 supported this work.
Footnotes
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
Data deposition: The electron microscopy density maps reported in this paper have been deposited in the Electron Microscopy Data Bank, http://www.emdatabank.org/ (accession no(s). EMD-5936–EMD-5939).
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1317972111/-/DCSupplemental.
References
- 1.Hunter E. Viral entry and receptors. In: Coffin JM, Hughes SH, Varmus HE, editors. Retroviruses. Cold Spring Harbor, NY: Cold Spring Harbor Lab Press; 1997. pp. 71–119. [PubMed] [Google Scholar]
- 2.Hallenberger S, et al. Inhibition of furin-mediated cleavage activation of HIV-1 glycoprotein gp160. Nature. 1992;360(6402):358–361. doi: 10.1038/360358a0. [DOI] [PubMed] [Google Scholar]
- 3.Fass D, Harrison SC, Kim PS. Retrovirus envelope domain at 1.7 angstrom resolution. Nat Struct Biol. 1996;3(5):465–469. doi: 10.1038/nsb0596-465. [DOI] [PubMed] [Google Scholar]
- 4.Kobe B, Center RJ, Kemp BE, Poumbourios P. Crystal structure of human T cell leukemia virus type 1 gp21 ectodomain crystallized as a maltose-binding protein chimera reveals structural evolution of retroviral transmembrane proteins. Proc Natl Acad Sci USA. 1999;96(8):4319–4324. doi: 10.1073/pnas.96.8.4319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wilson IA, Skehel JJ, Wiley DC. Structure of the haemagglutinin membrane glycoprotein of influenza virus at 3 Å resolution. Nature. 1981;289(5796):366–373. doi: 10.1038/289366a0. [DOI] [PubMed] [Google Scholar]
- 6.Chen J, et al. Structure of the hemagglutinin precursor cleavage site, a determinant of influenza pathogenicity and the origin of the labile conformation. Cell. 1998;95(3):409–417. doi: 10.1016/s0092-8674(00)81771-7. [DOI] [PubMed] [Google Scholar]
- 7.Welch BD, et al. Structure of the cleavage-activated prefusion form of the parainfluenza virus 5 fusion protein. Proc Natl Acad Sci USA. 2012;109(41):16672–16677. doi: 10.1073/pnas.1213802109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.McLellan JS, et al. Structure of RSV fusion glycoprotein trimer bound to a prefusion-specific neutralizing antibody. Science. 2013;340(6136):1113–1117. doi: 10.1126/science.1234914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pinter A, Honnen WJ, Tung JS, O’Donnell PV, Hämmerling U. Structural domains of endogenous murine leukemia virus gp70s containing specific antigenic determinants defined by monoclonal antibodies. Virology. 1982;116(2):499–516. doi: 10.1016/0042-6822(82)90143-x. [DOI] [PubMed] [Google Scholar]
- 10.Albritton LM, Tseng L, Scadden D, Cunningham JM. A putative murine ecotropic retrovirus receptor gene encodes a multiple membrane-spanning protein and confers susceptibility to virus infection. Cell. 1989;57(4):659–666. doi: 10.1016/0092-8674(89)90134-7. [DOI] [PubMed] [Google Scholar]
- 11.Wang H, Kavanaugh MP, North RA, Kabat D. Cell-surface receptor for ecotropic murine retroviruses is a basic amino-acid transporter. Nature. 1991;352(6337):729–731. doi: 10.1038/352729a0. [DOI] [PubMed] [Google Scholar]
- 12.Lavillette D, Boson B, Russell SJ, Cosset FL. Activation of membrane fusion by murine leukemia viruses is controlled in cis or in trans by interactions between the receptor-binding domain and a conserved disulfide loop of the carboxy terminus of the surface glycoprotein. J Virol. 2001;75(8):3685–3695. doi: 10.1128/JVI.75.8.3685-3695.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Barnett AL, Davey RA, Cunningham JM. Modular organization of the Friend murine leukemia virus envelope protein underlies the mechanism of infection. Proc Natl Acad Sci USA. 2001;98(7):4113–4118. doi: 10.1073/pnas.071432398. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pinter A, Kopelman R, Li Z, Kayman SC, Sanders DA. Localization of the labile disulfide bond between SU and TM of the murine leukemia virus envelope protein complex to a highly conserved CWLC motif in SU that resembles the active-site sequence of thiol-disulfide exchange enzymes. J Virol. 1997;71(10):8073–8077. doi: 10.1128/jvi.71.10.8073-8077.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Opstelten DJ, Wallin M, Garoff H. Moloney murine leukemia virus envelope protein subunits, gp70 and Pr15E, form a stable disulfide-linked complex. J Virol. 1998;72(8):6537–6545. doi: 10.1128/jvi.72.8.6537-6545.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wallin M, Ekström M, Garoff H. Isomerization of the intersubunit disulphide-bond in Env controls retrovirus fusion. EMBO J. 2004;23(1):54–65. doi: 10.1038/sj.emboj.7600012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Förster F, Medalia O, Zauberman N, Baumeister W, Fass D. Retrovirus envelope protein complex structure in situ studied by cryo-electron tomography. Proc Natl Acad Sci USA. 2005;102(13):4729–4734. doi: 10.1073/pnas.0409178102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wu SR, et al. Turning of the receptor-binding domains opens up the murine leukaemia virus Env for membrane fusion. EMBO J. 2008;27(20):2799–2808. doi: 10.1038/emboj.2008.187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Fass D, et al. Structure of a murine leukemia virus receptor-binding glycoprotein at 2.0 angstrom resolution. Science. 1997;277(5332):1662–1666. doi: 10.1126/science.277.5332.1662. [DOI] [PubMed] [Google Scholar]
- 20.Green N, et al. Sequence-specific antibodies show that maturation of Moloney leukemia virus envelope polyprotein involves removal of a COOH-terminal peptide. Proc Natl Acad Sci USA. 1981;78(10):6023–6027. doi: 10.1073/pnas.78.10.6023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Schultz A, Rein A. Maturation of murine leukemia virus env proteins in the absence of other viral proteins. Virology. 1985;145(2):335–339. doi: 10.1016/0042-6822(85)90168-0. [DOI] [PubMed] [Google Scholar]
- 22.Henderson LE, Sowder R, Copeland TD, Smythers G, Oroszlan S. Quantitative separation of murine leukemia virus proteins by reversed-phase high-pressure liquid chromatography reveals newly described gag and env cleavage products. J Virol. 1984;52(2):492–500. doi: 10.1128/jvi.52.2.492-500.1984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ragheb JA, Anderson WF. pH-independent murine leukemia virus ecotropic envelope-mediated cell fusion: Implications for the role of the R peptide and p12E TM in viral entry. J Virol. 1994;68(5):3220–3231. doi: 10.1128/jvi.68.5.3220-3231.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rein A, Mirro J, Haynes JG, Ernst SM, Nagashima K. Function of the cytoplasmic domain of a retroviral transmembrane protein: p15E-p2E cleavage activates the membrane fusion capability of the murine leukemia virus Env protein. J Virol. 1994;68(3):1773–1781. doi: 10.1128/jvi.68.3.1773-1781.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Löving R, Kronqvist M, Sjöberg M, Garoff H. Cooperative cleavage of the R peptide in the Env trimer of Moloney murine leukemia virus facilitates its maturation for fusion competence. J Virol. 2011;85(7):3262–3269. doi: 10.1128/JVI.02500-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Löving R, Wu SR, Sjöberg M, Lindqvist B, Garoff H. Maturation cleavage of the murine leukemia virus Env precursor separates the transmembrane subunits to prime it for receptor triggering. Proc Natl Acad Sci USA. 2012;109(20):7735–7740. doi: 10.1073/pnas.1118125109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Shapiro SZ, Strand M, August JT. High molecular weight precursor polypeptides to structural proteins of Rauscher murine leukemia virus. J Mol Biol. 1976;107(4):459–477. doi: 10.1016/s0022-2836(76)80078-2. [DOI] [PubMed] [Google Scholar]
- 28.Ng VL, Wood TG, Arlinghaus RB. Processing of the env gene products of Moloney murine leukaemia virus. J Gen Virol. 1982;59(Pt 2):329–343. doi: 10.1099/0022-1317-59-2-329. [DOI] [PubMed] [Google Scholar]
- 29.Pinter A, Honnen WJ. O-linked glycosylation of retroviral envelope gene products. J Virol. 1988;62(3):1016–1021. doi: 10.1128/jvi.62.3.1016-1021.1988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Geyer R, et al. Oligosaccharides at individual glycosylation sites in glycoprotein 71 of Friend murine leukemia virus. Eur J Biochem. 1990;187(1):95–110. doi: 10.1111/j.1432-1033.1990.tb15281.x. [DOI] [PubMed] [Google Scholar]
- 31.Apte S, Sanders DA. Effects of retroviral envelope-protein cleavage upon trafficking, incorporation, and membrane fusion. Virology. 2010;405(1):214–224. doi: 10.1016/j.virol.2010.06.004. [DOI] [PubMed] [Google Scholar]
- 32.Sjöberg M, Wallin M, Lindqvist B, Garoff H. Furin cleavage potentiates the membrane fusion-controlling intersubunit disulfide bond isomerization activity of leukemia virus Env. J Virol. 2006;80(11):5540–5551. doi: 10.1128/JVI.01851-05. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Evans LH, Morrison RP, Malik FG, Portis J, Britt WJ. A neutralizable epitope common to the envelope glycoproteins of ecotropic, polytropic, xenotropic, and amphotropic murine leukemia viruses. J Virol. 1990;64(12):6176–6183. doi: 10.1128/jvi.64.12.6176-6183.1990. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Burkhart MD, D’Agostino P, Kayman SC, Pinter A. Involvement of the C-terminal disulfide-bonded loop of murine leukemia virus SU protein in a postbinding step critical for viral entry. J Virol. 2005;79(12):7868–7876. doi: 10.1128/JVI.79.12.7868-7876.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Burkhart MD, Kayman SC, He Y, Pinter A. Distinct mechanisms of neutralization by monoclonal antibodies specific for sites in the N-terminal or C-terminal domain of murine leukemia virus SU. J Virol. 2003;77(7):3993–4003. doi: 10.1128/JVI.77.7.3993-4003.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Zavorotinskaya T, Albritton LM. Failure To cleave murine leukemia virus envelope protein does not preclude its incorporation in virions and productive virus-receptor interaction. J Virol. 1999;73(7):5621–5629. doi: 10.1128/jvi.73.7.5621-5629.1999. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Fass D, Kim PS. Dissection of a retrovirus envelope protein reveals structural similarity to influenza hemagglutinin. Curr Biol. 1995;5(12):1377–1383. doi: 10.1016/s0960-9822(95)00275-2. [DOI] [PubMed] [Google Scholar]
- 38.Felkner RH, Roth MJ. Mutational analysis of the N-linked glycosylation sites of the SU envelope protein of Moloney murine leukemia virus. J Virol. 1992;66(7):4258–4264. doi: 10.1128/jvi.66.7.4258-4264.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Daniels R, Kurowski B, Johnson AE, Hebert DN. N-linked glycans direct the cotranslational folding pathway of influenza hemagglutinin. Mol Cell. 2003;11(1):79–90. doi: 10.1016/s1097-2765(02)00821-3. [DOI] [PubMed] [Google Scholar]
- 40.Kayman SC, Kopelman R, Projan S, Kinney DM, Pinter A. Mutational analysis of N-linked glycosylation sites of Friend murine leukemia virus envelope protein. J Virol. 1991;65(10):5323–5332. doi: 10.1128/jvi.65.10.5323-5332.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Li Z, Pinter A, Kayman SC. The critical N-linked glycan of murine leukemia virus envelope protein promotes both folding of the C-terminal domains of the precursor polyprotein and stability of the postcleavage envelope complex. J Virol. 1997;71(9):7012–7019. doi: 10.1128/jvi.71.9.7012-7019.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Liu J, Bartesaghi A, Borgnia MJ, Sapiro G, Subramaniam S. Molecular architecture of native HIV-1 gp120 trimers. Nature. 2008;455(7209):109–113. doi: 10.1038/nature07159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Mao Y, et al. Subunit organization of the membrane-bound HIV-1 envelope glycoprotein trimer. Nat Struct Mol Biol. 2012;19(9):893–899. doi: 10.1038/nsmb.2351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Li K, et al. Intersubunit disulfide isomerization controls membrane fusion of human T-cell leukemia virus Env. J Virol. 2008;82(14):7135–7143. doi: 10.1128/JVI.00448-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.





